Longitudinal Assessment of Ketamine-Assisted Community of Practice Therapy: A Retrospective 5-Year Study on Depression, Anxiety and PTSD
A. Kuzma-Hunt, Z. Hamadeh, V. TsangIntroduction
Ketamine-assisted therapy shows rapid benefit in treatment-resistant depression, anxiety, and PTSD (Walsh et al. BJPsych Open 2022;8:e19). The Roots to Thrive Ketamine assisted Therapy (RTT-KaT) program is a unique 12-week program that combines 12 Community of Practice (group therapy) sessions and three ketamine medicine sessions (Tsang et al. Ther Adv Psychopharmacol 2023;13:1–14). Predictors of treatment responses within this model are unclear.
Objectives
Evaluate pre–post changes in PHQ-9, GAD-7, PCL-5, and B-IPF; Assess whether baseline severity predicts improvement; Examine effects of concurrent pharmacotherapy or prior psychotherapy; Test whether Adverse Childhood Experiences (ACEs) predict treatment outcomes.
Methods
We conducted a retrospective cohort analysis of 305 participants enrolled in the 12-week RTT-KAT program (2018–2023), using validated scales (PHQ-9, GAD-7, PCL-5, BIPF) administered at baseline and post-program (Figure 1). Paired t-tests with 95% CI and Cohen’s d. Linear regressions assessed whether baseline severity predicted improvement, while ANCOVA of post-scores adjusted for baseline and pharmacotherapy/psychotherapy status. ACEs were evaluated using Pearson correlation and Welch’s t-tests (ACE ≥4 vs <4). Multiple analyses accounted for multiple comparisons with α=0.05.
Results
Significant reductions were observed for PHQ-9, GAD-7, PCL-5 (all
p
<0.001,
d
>0.8) and moderate improvement for B-IPF (
d
=0.5; Figure 1). Higher baseline severity predicted greater absolute improvement across scales (negative β,
p
<0.001; Figure 2). After adjusting for baseline severity, concurrent pharmacotherapy and prior psychotherapy showed no effect on depression or anxiety. Pharmacotherapy predicted higher adjusted PTSD post-scores (β=4.48, 95% CI 0.72–8.24,
p
=0.020) and a non-significant trend toward poorer functioning (
p
=0.064). ACEs showed no significant correlation with treatment response (|r|<0.08, all
p
>0.20). Threshold comparisons (ACE ≥4 vs <4) also showed no group differences (all
p
>0.13). All confidence intervals included zero, indicating no evidence of either linear or threshold associations. Substance-use history was unrelated to outcomes (all p>0.20). Similarly, current alcohol, cannabis, or tobacco use did not significantly affect post-treatment outcomes once baseline severity was controlled.
Image 1: Long description.